US10594943B2 - Camera calibration device and camera calibration system - Google Patents

Camera calibration device and camera calibration system Download PDF

Info

Publication number
US10594943B2
US10594943B2 US15/306,672 US201515306672A US10594943B2 US 10594943 B2 US10594943 B2 US 10594943B2 US 201515306672 A US201515306672 A US 201515306672A US 10594943 B2 US10594943 B2 US 10594943B2
Authority
US
United States
Prior art keywords
vehicle
camera
calibration
video
cameras
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active, expires
Application number
US15/306,672
Other languages
English (en)
Other versions
US20170061622A1 (en
Inventor
Morihiko SAKANO
Keiji Sato
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Faurecia Clarion Electronics Co Ltd
Original Assignee
Clarion Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Clarion Co Ltd filed Critical Clarion Co Ltd
Assigned to CLARION CO., LTD. reassignment CLARION CO., LTD. ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS). Assignors: SAKANO, MORIHIKO, SATO, KEIJI
Publication of US20170061622A1 publication Critical patent/US20170061622A1/en
Application granted granted Critical
Publication of US10594943B2 publication Critical patent/US10594943B2/en
Active legal-status Critical Current
Adjusted expiration legal-status Critical

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N7/00—Television systems
    • H04N7/18—Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast
    • H04N7/181—Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast for receiving images from a plurality of remote sources
    • H04N5/23296—
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60R—VEHICLES, VEHICLE FITTINGS, OR VEHICLE PARTS, NOT OTHERWISE PROVIDED FOR
    • B60R1/00—Optical viewing arrangements; Real-time viewing arrangements for drivers or passengers using optical image capturing systems, e.g. cameras or video systems specially adapted for use in or on vehicles
    • G06K9/00744—
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T11/00—2D [Two Dimensional] image generation
    • G06T11/60—Editing figures and text; Combining figures or text
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/80—Analysis of captured images to determine intrinsic or extrinsic camera parameters, i.e. camera calibration
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/40—Scenes; Scene-specific elements in video content
    • G06V20/46—Extracting features or characteristics from the video content, e.g. video fingerprints, representative shots or key frames
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N17/00—Diagnosis, testing or measuring for television systems or their details
    • H04N17/002—Diagnosis, testing or measuring for television systems or their details for television cameras
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
    • H04N23/57—Mechanical or electrical details of cameras or camera modules specially adapted for being embedded in other devices
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
    • H04N23/60—Control of cameras or camera modules
    • H04N23/69—Control of means for changing angle of the field of view, e.g. optical zoom objectives or electronic zooming
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60R—VEHICLES, VEHICLE FITTINGS, OR VEHICLE PARTS, NOT OTHERWISE PROVIDED FOR
    • B60R2300/00—Details of viewing arrangements using cameras and displays, specially adapted for use in a vehicle
    • B60R2300/10—Details of viewing arrangements using cameras and displays, specially adapted for use in a vehicle characterised by the type of camera system used
    • B60R2300/105—Details of viewing arrangements using cameras and displays, specially adapted for use in a vehicle characterised by the type of camera system used using multiple cameras
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60R—VEHICLES, VEHICLE FITTINGS, OR VEHICLE PARTS, NOT OTHERWISE PROVIDED FOR
    • B60R2300/00—Details of viewing arrangements using cameras and displays, specially adapted for use in a vehicle
    • B60R2300/40—Details of viewing arrangements using cameras and displays, specially adapted for use in a vehicle characterised by the details of the power supply or the coupling to vehicle components
    • B60R2300/402—Image calibration
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60R—VEHICLES, VEHICLE FITTINGS, OR VEHICLE PARTS, NOT OTHERWISE PROVIDED FOR
    • B60R2300/00—Details of viewing arrangements using cameras and displays, specially adapted for use in a vehicle
    • B60R2300/60—Details of viewing arrangements using cameras and displays, specially adapted for use in a vehicle characterised by monitoring and displaying vehicle exterior scenes from a transformed perspective
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10016—Video; Image sequence
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20212—Image combination
    • G06T2207/20221—Image fusion; Image merging
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30244—Camera pose
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30248—Vehicle exterior or interior
    • G06T2207/30252—Vehicle exterior; Vicinity of vehicle
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30248—Vehicle exterior or interior
    • G06T2207/30252—Vehicle exterior; Vicinity of vehicle
    • G06T2207/30256—Lane; Road marking
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/40—Extraction of image or video features
    • G06V10/48—Extraction of image or video features by mapping characteristic values of the pattern into a parameter space, e.g. Hough transformation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/50—Context or environment of the image
    • G06V20/52—Surveillance or monitoring of activities, e.g. for recognising suspicious objects
    • G06V20/54—Surveillance or monitoring of activities, e.g. for recognising suspicious objects of traffic, e.g. cars on the road, trains or boats
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/50—Context or environment of the image
    • G06V20/56—Context or environment of the image exterior to a vehicle by using sensors mounted on the vehicle
    • G06V20/588—Recognition of the road, e.g. of lane markings; Recognition of the vehicle driving pattern in relation to the road
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
    • H04N23/60—Control of cameras or camera modules
    • H04N23/698—Control of cameras or camera modules for achieving an enlarged field of view, e.g. panoramic image capture
    • H04N5/23238—

Definitions

  • the present invention relates to a camera calibration device and a camera calibration system.
  • camera-specific information such as an optical characteristic (a focal distance and a lens distortion) and a size of an image pickup device, and information (external parameter) on an attached position and an angle of each camera are required.
  • the videos captured by the cameras are transformed in overview using camera parameters obtained by synthesizing the inner parameter and the external parameter as described above.
  • a video captured from the overview point can be virtually generated by synthesizing the overview videos obtained from the videos.
  • the cameras are attached to the vehicle at positions and with angles in conformity to design values. At that time, there occurs an error inevitably.
  • the overview video is generated on the basis of the design values regardless of such an error
  • the overview video captured from the overview point is necessarily not an expected one, and a deviation occurs in the video by an amount of the error.
  • an influence of the deviation remarkably appears in a boundary area of images of the cameras in the synthesized image, which is greater on appearance than a case where a single camera is used.
  • a correction (called calibration) of the error caused from the design value of the camera parameter is performed.
  • a method of estimating a current installation state of the camera from the captured video is employed in place of a method of mechanically adjusting the installation state thereof.
  • a typical method of estimating the installation state from the captured video there is typically employed a method of accurately providing a pattern (a calibration chart) printed in a sheet or a plate at a predetermined position, and correcting the camera parameter such that the actually captured video is matched to a video captured by a camera which is manufactured and provided in conformity to the design value.
  • the attachment state of the camera is not adjusted, but numerical values in a program related to the attachment position and the angle of the camera are corrected through an image transformation.
  • the calibration is executed by simulating an empty condition of no one passenger or a specific loading state such as a case where a driver sits in a driver seat. Therefore, the deviation is not generated in the video in the same state as the actual calibration such as the empty state or the specific loading state.
  • PTLs 1 and 2 disclose technologies of correcting the camera parameter in running of the vehicle.
  • An online calibration method of a vehicle camera disclosed in PTL 1 is a method in which an adjacent portion of a road is captured by at least two cameras, a road characteristic in a longitudinal direction is specified in an image frame, a road characteristic in the longitudinal direction specified in at least two image frames captured by two camera is selected such that two image frames are matched by a single line therebetween, a matching rate of the single line is analyzed to determine an offset of the line between two image frames, and the offset of the determined line is applied for the calibration of the camera.
  • an online calibration method of a vehicle camera disclosed in PTL 2 is a method in which a part of a road is captured by the camera, a road characteristic in the longitudinal direction is specified in an image frame, a point along the specified road characteristic in the longitudinal direction is extracted and the extracted point is transformed into a virtual road plane by a perspective mapping in consideration of a given camera parameter, the extracted point thus transformed is analyzed with respect to the vehicle to determine a deviation from a line in parallel to the vehicle of the point, and the measured deviation is used to define an offset correction of the camera.
  • the calibration is realized using the parallelism of the vehicle with respect to a white line in the online calibration method of the vehicle camera disclosed in PTL 2, there is a problem in that the calibration can be executed only in a case where the vehicle is in parallel to the white line.
  • the calibration is executed only in a case where the vehicle runs at a certain speed (50 km) or more in order to secure the parallelism of the vehicle with respect to the white line. For example, in a case where the vehicle runs at a low speed in a street, there is also a problem in that the calibration is not possible to execute.
  • the invention has been made in view of the problems, and an object thereof is to provide a camera calibration device and a camera calibration system which can execute the calibration even under a situation that the loading state of the vehicle changes, and can execute the calibration by accurately estimating all the camera parameters without using the parallelism of the vehicle with respect to the white line for example.
  • a camera calibration device is a camera calibration device that is mounted in a vehicle and executes calibration on a plurality of cameras capturing an ambient environment of the vehicle, including: a video acquisition unit that acquires a video captured by the camera; a feature extraction unit that extracts a predetermined feature quantity from the video; a posture estimation unit that estimates a posture of the vehicle on the basis of the predetermined feature quantity; a translation correction unit that corrects a position in a translation direction of the camera with respect to a ground surface on the basis of information obtained from the calibration executed in the past; and a camera parameter calculation unit that calculates a camera parameter related to a posture of the camera with respect to the ground surface on the basis of the posture of the vehicle and a position of the camera in the translation direction with respect to the ground surface.
  • a camera calibration system acoustic to the present invention includes: the camera calibration device; the plurality of cameras that are mounted in the vehicle such that the captured videos are partially overlapped or adjacent; a synthesized video generation device that corrects the videos captured by the plurality of cameras using camera parameters of the cameras obtained from the camera calibration device so as to generate a synthesized video; and a display device that displays the synthesized video.
  • the calibration can be executed even under a situation that the loading state of the vehicle changes, and can be executed by accurately estimating all the camera parameters without using the parallelism of the vehicle with respect to the white line for example. Therefore, it is possible to generate an overview video with high accuracy according to a loading state of the vehicle under any situation.
  • FIG. 1 is a diagram illustrating the entire configuration of a camera calibration system according to an embodiment of the invention.
  • FIG. 2 is a flowchart for describing a procedure until calibration is executed at the time of product shipment and at the time of usage.
  • FIG. 3 is a diagram illustrating an example of an overview video before the calibration at the time of usage is executed.
  • FIG. 4 is a diagram illustrating an example of the overview video in process of executing the calibration at the time of usage.
  • FIG. 5 is a diagram illustrating an example of the overview video after the calibration at the time of usage is executed.
  • FIG. 6 is a diagram illustrating an example of an inner configuration of the camera calibration device illustrated in FIG. 1 .
  • FIG. 7 is a flowchart for describing a calibration process of the camera calibration device illustrated in FIG. 6 .
  • FIG. 8 is a diagram illustrating another example of the inner configuration of the camera calibration device illustrated in FIG. 1 .
  • FIG. 9 is a flowchart for describing the calibration process of the camera calibration device illustrated in FIG. 8 .
  • FIG. 10 is a diagram illustrating still another example of the inner configuration of the camera calibration device illustrated in FIG. 1 .
  • FIG. 11 is a flowchart for describing the calibration process of the camera calibration device illustrated in FIG. 10 .
  • FIG. 1 illustrates a system configuration for realizing an embodiment of a camera calibration system according to the invention.
  • a camera calibration system 100 in the drawing mainly includes four cameras 111 to 114 , a calculation device 101 , a RAM 102 , a ROM 103 , a display device 104 , a speed sensor 105 , a steering sensor 106 , a yaw rate sensor 107 , an input device 108 , and a communication device 109 .
  • the cameras 111 to 114 are mounted in a vehicle 1 , and provided on front, rear, right, and left sides of the vehicle 1 for example.
  • the cameras provided on the front and rear sides are attached to a vehicle body in the vicinity of a number plate, and the cameras provided on the right and left sides are attached to lower portions of side mirrors.
  • the camera 111 is attached on the front side of the vehicle 1 , the camera 112 on the rear side of the vehicle 1 , the camera 113 on the left side of the vehicle 1 , and the camera 114 on the right side of the vehicle 1 (see FIG. 3 ).
  • the cameras 111 to 114 are attached such that an optical axis faces from a horizontal direction to a perpendicular direction with respect to a horizontal plane parallel to the ground.
  • the cameras 111 to 114 are attached in accordance with known design information which is set in advance. However, actually there is an error in attachment, and such an error is unknown.
  • a fish eye camera having a wide angle is employed as each of the cameras 111 to 114 to acquire the video all around the vehicle 1 . Since the fish eye camera acquires a wide-angle video, the image is distorted on the basis of a known distortion function.
  • the four videos captured by the cameras 111 to 114 are transmitted to the calculation device 101 .
  • the speed sensor 105 , the steering sensor 106 , and the yaw rate sensor 107 are sensors for measuring speed, steering, and yaw rate. Sensor information measured by each sensor is transmitted to the calculation device 101 , and used in a calculation process of the calculation device 101 .
  • the input device 108 is a device such as a switch and a button which receives a user's operation, and is used in turning on/off a calibration function, initializing a calibration result, and changing a calibration method.
  • Various types of information input to the input device 108 through a user's operation are transmitted to the calculation device 101 .
  • the communication device 109 is a device which is used in communication with an external machine (not illustrated).
  • the calculation device 101 receives various types of information from the external machine through the communication device 109 , and transmits various types of information calculated by the calculation device 101 to the external machine.
  • Numerical data required in the calculation process of the calculation device 101 , and a variable of a program with respect to a processing result obtained in the middle of the calculation process are written in the RAM 102 .
  • the written data is appropriately read as needed in the calculation process of the calculation device 101 to be used in the calculation process.
  • video data captured by the cameras 111 to 114 is also stored in the RAM 102 .
  • a program for performing the calibration and information to be used without being rewritten among the information required in the program are stored in advance.
  • camera parameters such as a design value of an installation position and an angle of each camera (external parameter), a focal distance of each camera, a pixel size, a center of the optical axis, and the distortion function (inner parameters) are stored.
  • the calculation device 101 is a device which receives various types of information transmitted from the cameras 111 to 114 , the speed sensor 105 , the steering sensor 106 , the yaw rate sensor 107 , the input device 108 , and the communication device 109 , and executes the calculation process on the basis of a program.
  • the calculation device 101 executes a calculation process in which the videos input from the cameras 111 to 114 are transformed in viewpoint and synthesized to generate a video (overview video) viewed from the above.
  • the distortion of the videos captured by the fish eye cameras 111 to 114 is removed using the known distortion function which is stored in advance in the ROM 103 .
  • the videos obtained by removing the distortion are transformed in viewpoint and synthesized to obtain a video viewed from an overview point on the basis of the known design value related to the camera attachment which is stored in advance in the ROM 103 (a synthesized video generation device 115 ).
  • a viewpoint transformation/synthesis process may be realized by calculating a specific image of the overview video and specific images of the cameras 111 to 114 corresponding thereto using a well-known geometric transformation formula of the camera, and by assigning a luminance value of the image to the pixel of the overview video.
  • the corresponding pixel includes a decimal point and there is no subject pixel, there is performed a process of assigning an intermediate luminance value of the surrounding pixels through a well-known luminance interpolation processing.
  • the calculation device 101 performs the calculation process using the output results of the speed sensor 105 , the steering sensor 106 , the yaw rate sensor 107 , and the communication device 109 , or a process of switching the operation programs according to the input result of the input device 108 .
  • calculation device 101 is embedded with a camera calibration device 116 which executes calibration (correction) of the camera such that the overview video generated by the overview transformation/synthesis process becomes a video of the vehicle 1 viewed from right overhead.
  • the display device 104 receives the process result of the calculation device 101 , and presents the process result to the user using a display. For example, four videos of the cameras 111 to 114 are subjected to the viewpoint transformation/synthesis to generate the overview video, and displayed to a driver. In addition, the display device 104 can switch displaying contents according to the output of the calculation device 101 (for example, displaying only a video of the camera 112 which captures the rear side of the vehicle 1 ).
  • FIG. 2 is a flowchart for describing a procedure until the calibration is executed at the time of product shipment and at the time of usage.
  • calibration S 205 at the time of product shipment is executed, for example, after a camera attachment S 201 , a riding state reproduction S 202 , a position adjustment S 203 , and a calibration chart capturing S 204 .
  • the cameras 111 to 114 are attached to the vehicle body in the camera attachment S 201 , and a state where a passenger rides in the vehicle 1 is reproduced in the riding state reproduction S 202 .
  • the reason for executing the riding state reproduction S 202 is because a posture of the vehicle 1 is changed by a riding state, and angles of the cameras 111 to 114 with respect to the ground surface vary as the posture changes.
  • the vehicle 1 and the calibration chart are adjusted to satisfy a predetermined positional relation in the position adjustment S 203 .
  • the positional relation between the calibration chart and the cameras is determined to a defined positional relation, and then the calibration is executed.
  • the calibration chart is captured by the cameras 111 to 114 attached to the vehicle 1 in the calibration chart capturing S 204 .
  • the calibration is executed using the videos captured by the cameras 111 to 114 in the calibration S 205 at the time of product shipment.
  • the calibration is executed by a well-known technique.
  • the calibration is executed by a well-known technique.
  • the calibration can be executed even when the positional relation between the calibration chart and the vehicle 1 is undefined. In this case, a procedure of the position adjustment S 203 becomes unnecessary.
  • the calibration is executed on an assumption of a specific riding state as executed in the riding state reproduction S 202 .
  • the riding state is different from that at the time of calibration executed at the time of product shipment, there causes a deviation in a video boundary in the overview video. Therefore, it is not sufficient that the calibration is executed only in a specific riding state. For example, there is a need to execute the calibration at every time according to the riding state. Therefore, even at the time of usage including running and parking after the product shipment, the calibration in accordance with the riding state at that time is executed. In other words, as illustrated in FIG. 2 , calibration S 206 at the time of usage is executed after the calibration S 205 at the time of product shipment is executed.
  • the camera calibration device 116 built in the calculation device 101 executes the calibration at every time according to the riding state using information of ambient environments which can be captured by the cameras 111 to 114 at the time of usage of the vehicle 1 such as running and parking.
  • This embodiment is based on an assumption that the calibration is once executed in a factory, for example, by the procedure illustrated in FIG. 2 .
  • the camera parameters with respect to all the cameras 111 to 114 can be estimated without requiring a parallelism of the vehicle with respect to a white line (that is, a relative relation between the white line and the vehicle).
  • the calibration since the calibration is executed at the time of factory shipment as well as quality verification of the video captured by the camera, it is considered that an obtainable merit is large even on an assumption that the calibration is once executed as a constrain condition.
  • the calibration S 206 at the time of usage executed by the camera calibration device 116 will be described with reference to FIGS. 3 to 5 .
  • FIG. 3 illustrates an example of the overview video before the calibration at the time of usage is executed, in which an example of the overview video is illustrated when a loading state of the vehicle at the time of usage is changed from the loading state assumed in the calibration at the time of product shipment after the calibration is executed at the time of product shipment in the factory through the processes of S 201 to S 205 of FIG. 2 .
  • the white line of a road is drawn in the overview video
  • the loading state of the vehicle 1 is different from that assumed in advance. Therefore, the postures (angles and positions) of the cameras 111 to 114 with respect to the ground surface are changed, and the white lines are deviated in the video boundaries of the cameras 111 to 114 .
  • the calibration at the time of usage is executed in such a situation such that the overview video is corrected to eliminate the deviation of the white line.
  • the calibration is executed using the captured videos containing a linear feature quantity (feature quantity having linearity) in a longitudinal direction such as the white line in order to make the calibration executed in scenes as many as possible, and to make the calibration stable.
  • linear structures (the linear feature quantity, the white line in the drawing) are captured on both sides of the vehicle 1 .
  • two linear structures are necessarily captured by the camera 111 , two linear structures by the camera 112 , one linear structure by the camera 113 , and one linear structure by the camera 114 .
  • two linear structures captured by the cameras 111 and 112 are necessarily in parallel to each other.
  • the relative relation between the vehicle 1 and the linear structure is not required such as a relative angle between the vehicle 1 and the linear structure and a distance up to the linear structure.
  • the posture of the vehicle 1 (that is, the parameter indicating the posture of the vehicle 1 ) is estimated without directly estimating the parameter of the cameras 111 to 114 .
  • the camera parameters of the cameras 111 to 114 are estimated from the parameters indicating the posture of the vehicle 1 to execute the calibration.
  • the step of estimating the posture of the vehicle 1 is divided into a step of estimating a pitch angle of the vehicle 1 and a step of estimating a roll angle and a height of the vehicle 1 .
  • the pitch angle of the vehicle 1 is estimated on the basis of the parallelism between parallel straight lines captured by the cameras 111 and 112 installed on the front and rear sides of the vehicle 1 .
  • an overview video as illustrated in FIG. 4 is obtained.
  • the camera parameter is corrected such that the parallel straight lines captured by the cameras 111 and 112 are in parallel on the overview video.
  • the roll angle and the height of the vehicle 1 are estimated such that a deviation of the straight line in the video boundary of the overview video is eliminated.
  • the roll angle and the height of the vehicle 1 are estimated to correct the camera parameter, and the camera parameters of all the cameras are estimated in accordance with a variation of the vehicle posture to execute the calibration.
  • the overview video having no deviation in the video boundary is obtained as illustrated in FIG. 5 .
  • FIG. 6 is a diagram illustrating an example (first embodiment) of an inner configuration of the camera calibration device illustrated in FIG. 1 , in which an inner configuration of the camera calibration device to realize the calibration at the time of usage is illustrated.
  • FIG. 7 is a flowchart for describing a calibration process of the camera calibration device illustrated in FIG. 6 . Further, the calibration process of the camera calibration device illustrated in FIG. 6 is executed by loading a program stored in advance in the ROM 103 .
  • the camera calibration device 116 illustrated in FIG. 6 mainly includes a calibration execution determination unit 201 , a video acquisition unit 202 , a feature extraction unit 203 , and a calibration unit 204 .
  • the calibration execution determination unit 201 of the camera calibration device 116 determines whether the calibration is necessary for the camera (S 701 of FIG. 7 ). For example, the scenes showing the white lines in all four videos of the overview video captured by the cameras 111 to 114 are determined whether there occurs a deviation in the white lines in the video boundaries of the overview video. It is automatically determined whether the calibration is necessary for the camera.
  • the white line from the video is recognized using a well-known technique, a position in the video is calculated, a deviation amount at a predetermined video boundary position is measured, and it is determined whether the deviation amount exceeds a predetermined threshold so as to determine whether there occurs the deviation in the white line.
  • a method of determining whether the calibration is necessary by detecting a change in the loading state of the vehicle using a sensor (a gyro sensor) which directly estimates a posture change of the vehicle.
  • a method of determining whether the calibration of the camera is necessary using the linear feature quantity (for example, information on a white line, a curb stone, and a road end boundary including a broken line which are temporally stored in the RAM 102 ) extracted by the feature extraction unit 203 described below, and using the parallelism of the linear feature quantities in parallel to each other on both sides of (around) the vehicle in the video together with the deviation amount of the linear feature quantity in each video boundary in the overview video.
  • the video acquisition unit 202 acquires the videos captured by four cameras 111 to 114 attached to the vehicle 1 from the RAM 102 (S 702 of FIG. 7 ). Further, when there is a deviation in synchronization in the videos captured by the cameras 111 to 114 , the deviation appears as an error in the calibration. Therefore, it is desirable that the videos captured by the cameras 111 to 114 be stored in the RAM 102 in a perfect synchronization, or a deviation time of the video can be acquired. Further, the video acquisition unit 202 may directly acquire the videos from the cameras 111 to 114 .
  • the feature extraction unit 203 extracts a predetermined feature quantity to be used in the calibration, particularly the linear feature quantity (for example, a white line, a curb stone, and a road end boundary including a broken line) in the longitudinal direction (that is, a front and rear direction of the vehicle) in the videos captured by the cameras 111 to 114 acquired by the video acquisition unit 202 (S 703 of FIG. 7 ).
  • the linear feature quantity can be extracted by a method, for example, in which each of the cameras 111 to 114 generates a video with a distortion of the fish eye camera eliminated using the designed distortion function, a well-known edge extraction is executed in the video, and a well-known Hough transformation is used for an edge characteristic point.
  • the calibration unit 204 executes the calibration of the camera using the feature quantity obtained by the feature extraction unit 203 . While it is assumed that the calibration process in the calibration unit 204 is executed once in a factory, the parallelism of the vehicle with respect to the feature quantity in the longitudinal direction (that is, a relative relation between the feature quantity in the longitudinal direction and the vehicle) is not necessary, but all the camera parameters can be estimated only by the feature quantity in the longitudinal direction. In the calibration process, the camera parameters in the cameras 111 to 114 are not directly estimated, but the posture of the vehicle 1 is estimated. All the camera parameters are estimated through the posture of the vehicle 1 . Each camera is attached to a rigid vehicle body.
  • the angle and the position of each camera also vary in an interlocking manner. For example, when a passenger rides in a front seat of the vehicle 1 and the front side of the vehicle 1 goes down, the angle of the camera 111 attached on the front side of the vehicle 1 goes downward, and the camera 112 attached on the rear side of the vehicle 1 faces upward by the same angle. When a passenger rides in a right seat of the vehicle 1 and the right side of the vehicle 1 goes down, the camera 114 attached on the right side of the vehicle 1 faces downward, and the camera 113 attached on the left side of the vehicle 1 faces upward by the same angle.
  • the camera 111 attached on the front side of the vehicle 1 and the camera 112 attached on the rear side of the vehicle 1 are slightly rotated to the right side with respect to an optical axis.
  • the variation of each of the cameras 111 to 114 is interlocked with the variation of the posture of the vehicle body.
  • the variations of these cameras 111 to 114 are uniquely determined according to the variation of the posture of the vehicle body.
  • the camera parameters of the cameras 111 to 114 are not individually estimated, but the posture of the vehicle 1 is estimated and the camera parameters are calculated from the estimated posture of the vehicle 1 in the calibration process of the calibration unit 204 .
  • the roll angle of the camera 111 attached on the front side of the vehicle 1 can be obtained when the correction is performed such that the white line in the video is aligned in the just vertical direction.
  • the parameter of the vehicle posture can be estimated only by the straight line in the longitudinal direction in the video.
  • the positions and the angles of the cameras 111 to 114 attached to the vehicle 1 can be calculated, so that all the camera parameters can be estimated only by the straight line in the longitudinal direction in the video.
  • the calibration is necessarily executed in advance at the time of factory shipment.
  • the calibration is completed at the time of factory shipment, there is no deviation in the overview video in the loading state which is assumed at the time of the calibration.
  • the deviation in the overview video is caused by the change of the vehicle posture in accordance with the change of the loading state of the vehicle 1 . Therefore, when the variation of the camera posture according to the variation of the vehicle posture is canceled, it is possible to make the overview video have no deviation as it was. Therefore, it is possible to generate an overview video having no deviation by estimating the vehicle posture and by calculating the angle and the position of the camera according to the variation.
  • the overview video only returns to an uncalibrated state at the initial state and the deviation of the video is left as it is even when the angle and the position of the camera are corrected by a variation of the vehicle posture. Therefore, it is not possible to generate the overview video having no deviation. For this reason, the calibration is necessarily executed in the calibration process in advance.
  • the calibration unit 204 includes a posture estimation unit 301 , a translation correction unit 304 , and a camera parameter calculation unit 305 .
  • the posture estimation unit 301 is configured by a pitch angle estimation unit 302 and a roll angle/height/rotation center estimation unit 303 .
  • the pitch angle estimation unit 302 of the posture estimation unit 301 executes a pitch angle estimation of the vehicle 1 using the parallelism of the straight line (the linear feature quantity) in the longitudinal direction in the video captured by the cameras 111 and 112 attached on the front and rear sides of the vehicle 1 (S 704 of FIG. 7 ).
  • the parallel straight lines are displayed in parallel in an ideal overview video.
  • the parallel straight lines in the overview video are displayed in a “/ ⁇ ” shape which is not in parallel.
  • the pitch angle of the vehicle 1 is estimated such that the straight lines comes to be in parallel.
  • an inner product of a linear vector is set to an evaluation function on the basis of the parallelism of a straight line equation in each of the cameras 111 and 112 , and the pitch angle of the vehicle 1 is optimized such that the inner product in each of the cameras 111 and 112 approaches a value near “1” as close as possible.
  • the straight line (linear feature quantity) used herein is obtained by the feature extraction unit 203 described above. Further, the above optimization may be realized using a well-known technique such as a steepest descent method. In other words, for example, a process may be repeatedly executed in which an evaluation function related to the pitch angle of the vehicle 1 is obtained and the pitch angle of the vehicle 1 is slightly changed for the evaluation function to approach a target value.
  • the roll angle/height/rotation center estimation unit 303 of the posture estimation unit 301 estimates the roll angle of the vehicle 1 , the height of the vehicle 1 , and the rotation center of the vehicle 1 to eliminate the deviation of the straight line (linear feature quantity) in the longitudinal direction appearing in the video boundary of the overview video (S 705 of FIG. 7 ).
  • the pitch angle of the vehicle 1 is fixed to a pitch angle obtained by the pitch angle estimation unit 302 .
  • the evaluation function is designed to indicate the deviation of the straight line in the longitudinal direction appearing in the video boundary of the overview video, and the parameters related to the roll angle of the vehicle 1 , the height of the vehicle 1 , and the rotation center of the vehicle 1 are optimized to minimize the evaluation function.
  • the evaluation function is obtained by a total value obtained by calculating the deviation of the straight line in the video boundary of the overview video from the video boundary between the cameras 111 and 113 , the video boundary between the cameras 111 and 114 , the video boundary between the cameras 112 and 113 , and the video boundary between the cameras 112 and 114 .
  • the straight line (linear feature quantity) used herein is obtained by the feature extraction unit 203 described above. Further, while the above optimization is executed using a well-known technique, there are a plurality of parameters to be estimated. Therefore, it is desirable to employ a global optimizing technique in place of a gradient method such as the steepest descent method.
  • the process of slightly changing the roll angle of the vehicle 1 , the height of the vehicle 1 , and the rotation center of the vehicle 1 is repeatedly executed to minimize the evaluation function indicating the deviation of the straight line in the video boundary of the overview video. Further, a fact that the camera parameters may be different depending on the rotation center of the vehicle 1 is taken into consideration, and the optimization is executed including the rotation center of the vehicle 1 .
  • the translation correction unit 304 corrects a parameter (translation parameter) corresponding to the translation (movement in a direction parallel to the ground surface) of the vehicle 1 with respect to the ground surface (that is, a position (position in plan view) in a translation direction except the height of the camera) on the basis of the information obtained from the past calibration (S 706 of FIG. 7 ).
  • a parameter translation parameter
  • the position of the camera with respect to the ground surface is not possible to estimate since information indicating an absolute distance between the ground surface and the camera is not measured.
  • the deviation of the translation direction of the camera is extremely small in an actually obtainable variation range of the posture of the vehicle body. Therefore, for example, a value of the calibration executed at the time of product shipment which is stored in the ROM 103 is used.
  • a camera as a reference (reference camera) is selected (for example, the camera 111 ), and all the positions of the cameras 111 to 114 are translationally corrected to cancel an error of the translation direction between the calibration value at the time of product shipment of the camera 111 and the current value.
  • the camera parameter calculation unit 305 calculates the camera parameters related to the postures of the cameras 111 and 114 corresponding to the posture of the vehicle 1 from the posture of the vehicle 1 obtained by the posture estimation unit 301 and the parameter obtained by the translation correction unit 304 (S 707 of FIG. 7 ). Further, the camera parameters of the corresponding cameras 111 to 114 can be uniquely calculated by coordinate transformation.
  • the camera parameters of all the cameras can be delicately estimated only by the feature quantity in the longitudinal direction without necessitating the parallelism of the vehicle with respect to the feature quantity in the longitudinal direction (that is, a relative relation between the feature quantity in the longitudinal direction and the vehicle).
  • the camera parameter it is possible to generate the overview video with high accuracy in accordance with the loading state of the vehicle.
  • the calibration when the calibration is executed at the time of usage as described above, the calibration is desirably executed in a state (normal running state) where the vehicle runs on a straight lane in order to extract the linear feature quantity in the longitudinal direction from the video for example.
  • a state normal running state
  • FIG. 8 is a diagram illustrating another example (second embodiment) of an inner configuration of the camera calibration device illustrated in FIG. 11 .
  • FIG. 9 is a flowchart for describing the calibration process of the camera calibration device illustrated in FIG. 8 .
  • a camera calibration device 116 A illustrated in FIG. 8 is mainly different from the camera calibration device 116 illustrated in FIG. 6 in that a sensor information acquisition unit, a normal running determination unit, and a data availability determination unit are additionally provided, and other configurations are almost the same. Therefore, the description in the following will be made about only the configurations different from those of the camera calibration device 116 illustrated in FIG. 6 .
  • the same configurations as those of the camera calibration device 116 will be attached with the same symbols, and the detailed descriptions will be omitted.
  • the camera calibration device 116 A illustrated in FIG. 8 mainly includes a calibration execution determination unit 201 A, a video acquisition unit 202 A, a feature extraction unit 203 A, and a calibration unit 204 A. Further, a sensor information acquisition unit 205 A, a normal running state determination unit 206 A, and a data availability determination unit 207 A are also provided.
  • the calibration execution determination unit 201 A determines whether there is a need to execute the calibration of the camera (S 901 of FIG. 9 ). In a case where it is determined that there is a need to execute the calibration of the camera, the video acquisition unit 202 A acquires the videos captured by four cameras 111 to 114 attached to the vehicle 1 from the RAM 102 (S 902 of FIG. 9 ). Further, the sensor information acquisition unit 205 A acquires the sensor information of various types of sensors such as the speed sensor 105 , the steering sensor 106 , and the yaw rate sensor 107 (S 903 of FIG. 9 ). Herein, the sensor information is desirably acquired in synchronization with the video. Further, S 902 and S 903 illustrated in FIG. 9 may be executed at the same time, or S 902 may be executed after S 903 .
  • the normal running state determination unit 206 A determines whether the vehicle is in a normal running state (S 904 of FIG. 9 ).
  • the normal running state is a state in which there is no change in vehicle posture caused by accelerating/decelerating or turning, and the vehicle runs on the straight lane.
  • a reason for determining that there is no change in the vehicle posture caused by accelerating/decelerating or turning and the vehicle runs on the straight lane is that the camera parameters of the video having different vehicle posture are different and not able to be simultaneously used as the feature quantity at the time of executing the calibration, and another reason is to eliminate a video of a scene from which information necessary for the calibration in running curve is not acquired since the linear feature quantity (for example, a white line, a curb stone, and a road end boundary including a broken line) in the longitudinal direction is used as the feature quantity to be used in the vehicle posture estimation.
  • the linear feature quantity for example, a white line, a curb stone, and a road end boundary including a broken line
  • no accelerating/decelerating in the case of no change in the vehicle posture caused by accelerating/decelerating or turning can be determined from that a certain speed continues for a constant time period on the basis of the information acquired from the speed sensor 105 .
  • no change caused by turning can be determined from that an absolute value of the information acquired from the yaw rate sensor 107 is smaller than a predetermined threshold near to “0”.
  • a steering angle information acquired from the steering sensor 106
  • the feature extraction unit 203 A extracts a predetermined feature quantity to be used in the calibration from the videos captured by the cameras 111 to 114 (S 905 of FIG. 9 ).
  • the videos captured by the cameras 111 to 114 are discarded because the videos are not possible to use in the calibration.
  • a video of the next scene and the sensor information of various types of sensors are acquired.
  • the data availability determination unit 207 A determines whether the parallel white lines (linear feature quantities) available in the calibration are captured in the videos captured by the cameras 111 to 114 (S 906 of FIG. 9 ). Specifically, the data availability determination unit 207 A recognizes the white lines (linear feature quantities) captured in the cameras 111 to 114 as images, and determines whether both conditions are satisfied (the feature quantity extracted by the feature extraction unit 203 A is aligned in a straight line shape, and a difference between the video captured in the past and the white line is small). The determination on whether the feature quantity extracted by the feature extraction unit 203 A is aligned can be made by measuring a linearity of the white line. For example, each edge point of the white line is fitted to be aligned in a straight line.
  • a fitting error is equal to or less than a predetermined value
  • the white line is a straight line.
  • the fitting error is larger than a predetermined value
  • the white line is not a straight line.
  • the determination on whether a difference between the video captured in the past and the white line captured at the current time point is small can be made, for example, by calculating an angle of the white lines being paralleled similarly to the pitch angle estimation process of the vehicle.
  • an angular error with respect to an average angle calculated in the past is equal to or less than a predetermined value
  • it may be considered that the difference is small.
  • the angular error is larger than a predetermined value
  • the difference is large.
  • the calibration unit 204 A calculates the camera parameters of the cameras using the feature quantity obtained by the feature extraction unit 203 A as described above so as to execute the calibration (S 907 to S 910 of FIG. 9 ).
  • the videos captured by the cameras 111 to 114 are discarded because the videos are not possible to use in the calibration. A video of the next scene and the sensor information of various types of sensors are acquired.
  • the camera parameters of all the cameras can be more delicately and efficiently estimated only by the feature quantity in the longitudinal direction without necessitating the parallelism of the vehicle with respect to the feature quantity in the longitudinal direction (that is, a relative relation between the feature quantities in the longitudinal direction and the vehicle).
  • the camera parameter it is possible to generate the overview video with high accuracy in accordance with the loading state of the vehicle.
  • the calibration is desirably executed after a plurality of linear feature quantities captured at different positions in the video are collected.
  • FIG. 10 illustrates still another example (third embodiment) of an inner configuration of the camera calibration device illustrated in FIG. 1 .
  • FIG. 11 is a flowchart for describing the calibration process of the camera calibration device illustrated in FIG. 10 .
  • a camera calibration device 116 B of the embodiment illustrated in FIG. 10 is different from the camera calibration device 116 A illustrated FIG. 8 mainly in that a data accumulation determination unit is added, and other configurations are almost the same. Therefore, the description in the following will be made only about the configuration different from those of the camera calibration device 116 A illustrated in FIG. 8 .
  • the same configurations as those of the camera calibration device 116 A will be attached with the same symbols, and the detailed descriptions will be omitted.
  • the camera calibration device 116 B illustrated in FIG. 10 mainly includes a calibration execution determination unit 201 B, a video acquisition unit 202 B, a feature extraction unit 203 B, a calibration unit 204 B, a sensor information acquisition unit 205 B, a normal running state determination unit 206 B, and a data availability determination unit 207 B. Further, a data accumulation determination unit 208 B is also provided.
  • the calibration execution determination unit 201 B determines whether there is a need to execute the calibration of the camera (S 1101 of FIG. 11 ). In a case where it is determined that there is a need to execute the calibration of the camera, the video acquisition unit 202 B acquires the videos captured by four cameras 111 to 114 attached to the vehicle 1 from the RAM 102 (S 1102 of FIG. 11 ), and the sensor information acquisition unit 205 B acquires the sensor information of various types of sensors (S 1103 of FIG. 11 ).
  • the normal running state determination unit 206 B determines whether the vehicle is in the normal running state on the basis of the sensor information obtained by the sensor information acquisition unit 205 B (S 1104 of FIG. 11 ). In a case where it is determined that the vehicle is in the normal running state, the feature extraction unit 203 B extracts a predetermined feature quantity to be used in the calibration from the videos captured by the cameras 111 to 114 (S 1105 of FIG. 11 ). In addition, the data availability determination unit 207 B determines whether the extracted data is available to the calibration on the basis of the data extracted by the feature extraction unit 203 B ( 51106 of FIG. 11 ).
  • the extracted data is stored and accumulated in the RAM 102 serving as a data accumulation unit (S 1107 of FIG. 11 ).
  • the feature extraction unit 203 B stores a coefficient of an equation indicating a straight line obtained by the feature extraction unit 203 B.
  • the data accumulation determination unit 208 B determines whether an amount of data sufficient for the calibration is accumulated and the data accumulation is completed (S 1108 of FIG. 11 ).
  • the linear feature quantity (for example, a white line, a curb stone, and a road end boundary including a broken line) captured at the same position is meaningless as data for the cameras 111 to 114 attached to the vehicle 1 . Therefore, there is a need to collect a plurality of linear feature quantities captured at different positions (for example, the front, rear, right, and left sides around the vehicle 1 ).
  • the data accumulation determination unit 208 B sets in advance some areas for the cameras 111 to 114 , determines whether the straight line data of the areas is obtained for the entire areas without blind spots, and determines whether the data is accumulated as much amount as necessary for the calibration.
  • the calibration unit 204 B calculates the camera parameters of the cameras using the feature quantity accumulated in the RAM 102 serving as a data accumulation unit, and executes the calibration (S 1109 to S 1112 of FIG. 11 ).
  • the camera parameters of all the cameras can be more delicately and efficiently estimated only by the feature quantity in the longitudinal direction without necessitating the parallelism of the vehicle with respect to the feature quantity in the longitudinal direction (that is, a relative relation between the feature quantity in the longitudinal direction and the vehicle).
  • the camera parameter it is possible to generate the overview video with high accuracy in accordance with the loading state of the vehicle.
  • the predetermined feature quantity extracted from the video (in particular, the vehicle posture) is estimated on the basis of the linear feature quantity
  • the position of the camera in the translation direction with respect to the ground surface is corrected on the basis of the information obtained from the calibration executed in the past (for example, at the time of product shipment)
  • the camera parameter related to the posture of the camera with respect to the ground surface is calculated on the basis of the posture of the vehicle and the position of the camera in the translation direction with respect to the ground surface. Therefore, the calibration can be executed even under a situation where the loading state of the vehicle is changed.
  • all the camera parameters are estimated without providing a restriction, for example, the parallelism of the vehicle with respect to the white line, so that the calibration can be executed. Accordingly, it is possible to generate the overview video with high accuracy according to the loading state of the vehicle under any situation.
  • the invention is not limited to the above embodiments, and includes various modifications.
  • the above embodiments have been described in detail for easy understanding of the invention.
  • the invention is not necessarily to be provided with all the configurations described above.
  • some of the configurations of a certain embodiment may be replaced with the configurations of the other embodiments, and the configurations of the other embodiments may be added to the configurations of the subject embodiment.
  • some of the configurations of the embodiments may be omitted, replaced, and added to other configurations.
  • control lines and the information lines indicate something necessary for the description, and are not limited to those of all the control lines and the information lines necessary for a product. In practical, almost all the configurations may be considered to be connected to each other.

Landscapes

  • Engineering & Computer Science (AREA)
  • Multimedia (AREA)
  • Signal Processing (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Biomedical Technology (AREA)
  • Health & Medical Sciences (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • General Health & Medical Sciences (AREA)
  • Closed-Circuit Television Systems (AREA)
  • Studio Devices (AREA)
  • Image Processing (AREA)
  • Length Measuring Devices By Optical Means (AREA)
  • Mechanical Engineering (AREA)
  • Image Analysis (AREA)
US15/306,672 2014-09-30 2015-09-29 Camera calibration device and camera calibration system Active 2036-01-13 US10594943B2 (en)

Applications Claiming Priority (3)

Application Number Priority Date Filing Date Title
JP2014-201547 2014-09-30
JP2014201547A JP6371185B2 (ja) 2014-09-30 2014-09-30 カメラキャリブレーション装置及びカメラキャリブレーションシステム
PCT/JP2015/077517 WO2016052505A1 (ja) 2014-09-30 2015-09-29 カメラキャリブレーション装置及びカメラキャリブレーションシステム

Publications (2)

Publication Number Publication Date
US20170061622A1 US20170061622A1 (en) 2017-03-02
US10594943B2 true US10594943B2 (en) 2020-03-17

Family

ID=55630532

Family Applications (1)

Application Number Title Priority Date Filing Date
US15/306,672 Active 2036-01-13 US10594943B2 (en) 2014-09-30 2015-09-29 Camera calibration device and camera calibration system

Country Status (5)

Country Link
US (1) US10594943B2 (ja)
EP (1) EP3203724B1 (ja)
JP (1) JP6371185B2 (ja)
CN (1) CN106233711B (ja)
WO (1) WO2016052505A1 (ja)

Families Citing this family (27)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9884623B2 (en) * 2015-07-13 2018-02-06 GM Global Technology Operations LLC Method for image-based vehicle localization
US10558867B2 (en) * 2015-09-28 2020-02-11 Kyocera Corporation Image processing apparatus, stereo camera apparatus, vehicle, and image processing method
KR102457264B1 (ko) * 2016-05-02 2022-10-20 현대모비스 주식회사 어라운드 뷰 모니터링 시스템 및 이의 카메라 자동 공차 보정 방법
JP6614042B2 (ja) * 2016-06-15 2019-12-04 株式会社Ｊｖｃケンウッド 姿勢変化判定装置、俯瞰映像生成装置、俯瞰映像生成システム、姿勢変化判定方法およびプログラム
KR102528003B1 (ko) * 2016-07-27 2023-05-02 현대모비스 주식회사 차량용 카메라 공차 보정 방법 및 그 장치
JP2018157496A (ja) * 2017-03-21 2018-10-04 クラリオン株式会社 キャリブレーション装置
JP6724834B2 (ja) * 2017-03-23 2020-07-15 日産自動車株式会社 表示装置の制御方法および表示装置
JP6798926B2 (ja) * 2017-04-13 2020-12-09 クラリオン株式会社 車載カメラのキャリブレーション装置
JP6897442B2 (ja) 2017-09-12 2021-06-30 株式会社Ｊｖｃケンウッド 車両用装置、キャリブレーション結果判定システム、および、キャリブレーション結果判定方法、ならびに、プログラム
KR102326184B1 (ko) * 2017-11-27 2021-11-15 현대모비스 주식회사 후방카메라의 캘리브레이션 장치 및 방법
CN107945136B (zh) * 2017-12-04 2021-09-24 深圳市路畅科技股份有限公司 一种鱼眼图像矫正方法、系统、设备及计算机存储介质
JP7064163B2 (ja) * 2017-12-07 2022-05-10 コニカミノルタ株式会社 ３次元情報取得システム
JP7244488B2 (ja) * 2018-03-15 2023-03-22 株式会社村上開明堂 合成映像作成装置および合成映像作成方法並びに合成映像作成プログラム
JP2019180016A (ja) 2018-03-30 2019-10-17 パナソニックＩｐマネジメント株式会社 画像合成装置、画像合成方法及びプログラム
CN110570475A (zh) * 2018-06-05 2019-12-13 上海商汤智能科技有限公司 车载摄像头自标定方法及装置和车辆驾驶方法及装置
JP7314486B2 (ja) * 2018-09-06 2023-07-26 株式会社アイシン カメラキャリブレーション装置
JP7162489B2 (ja) * 2018-10-10 2022-10-28 フォルシアクラリオン・エレクトロニクス株式会社 キャリブレーション装置及びキャリブレーション方法
JP7148064B2 (ja) * 2018-10-25 2022-10-05 株式会社アイシン カメラパラメータ推定装置、カメラパラメータ推定方法、およびカメラパラメータ推定プログラム
JP7188228B2 (ja) * 2019-03-27 2022-12-13 株式会社デンソーテン 画像生成装置
US10997737B2 (en) * 2019-05-02 2021-05-04 GM Global Technology Operations LLC Method and system for aligning image data from a vehicle camera
JP7251310B2 (ja) * 2019-05-22 2023-04-04 株式会社Ｊｖｃケンウッド カメラ位置検出装置、カメラユニット、カメラ位置検出方法、及びプログラム
EP3991402A4 (en) 2019-06-25 2023-06-28 Snap Inc. Vanishing point stereoscopic image correction
DE102019211459B4 (de) * 2019-07-31 2021-02-18 Volkswagen Aktiengesellschaft Verfahren und Vorrichtung zum Überprüfen einer Kalibrierung von Umfeldsensoren
DE102020120873A1 (de) * 2019-08-12 2021-02-18 Motional AD LLC (n.d.Ges.d. Staates Delaware) Lokalisierung basierend auf vordefinierten merkmalen der umgebung
JP7065068B2 (ja) * 2019-12-13 2022-05-11 本田技研工業株式会社 車両周囲監視装置、車両、車両周囲監視方法およびプログラム
JP2023028939A (ja) * 2021-08-20 2023-03-03 本田技研工業株式会社 船舶用支援装置、船舶用支援方法、およびプログラム
CN116328278A (zh) * 2023-03-06 2023-06-27 浙江大沩人工智能科技有限公司 一种多人跑步的实时测速方法、系统、装置和介质

Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20080181591A1 (en) * 2007-01-26 2008-07-31 Sanyo Electric Co., Ltd. Camera posture estimation device, vehicle, and camera posture estimation method
US20120105642A1 (en) * 2009-06-29 2012-05-03 Panasonic Corporation Vehicle-mounted video display device
WO2012139660A1 (en) 2011-04-15 2012-10-18 Connaught Electronics Limited Online vehicle camera calibration based on road marking extractions
WO2012143036A1 (en) 2011-04-18 2012-10-26 Connaught Electronics Limited Online vehicle camera calibration based on continuity of features
WO2012145822A1 (en) 2011-04-25 2012-11-01 Magna International Inc. Method and system for dynamically calibrating vehicular cameras
US20130135474A1 (en) 2011-11-28 2013-05-30 Clarion Co., Ltd. Automotive Camera System and Its Calibration Method and Calibration Program
CN103686143A (zh) 2012-09-24 2014-03-26 歌乐株式会社 照相机的校准方法以及装置
WO2014054223A1 (ja) 2012-10-02 2014-04-10 株式会社デンソー キャリブレーション方法、及びキャリブレーション装置
JP2014074632A (ja) 2012-10-03 2014-04-24 Isuzu Motors Ltd 車載ステレオカメラの校正装置及び校正方法

Patent Citations (15)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20080181591A1 (en) * 2007-01-26 2008-07-31 Sanyo Electric Co., Ltd. Camera posture estimation device, vehicle, and camera posture estimation method
US20120105642A1 (en) * 2009-06-29 2012-05-03 Panasonic Corporation Vehicle-mounted video display device
WO2012139660A1 (en) 2011-04-15 2012-10-18 Connaught Electronics Limited Online vehicle camera calibration based on road marking extractions
WO2012143036A1 (en) 2011-04-18 2012-10-26 Connaught Electronics Limited Online vehicle camera calibration based on continuity of features
WO2012145822A1 (en) 2011-04-25 2012-11-01 Magna International Inc. Method and system for dynamically calibrating vehicular cameras
CN103136747A (zh) 2011-11-28 2013-06-05 歌乐株式会社 车载摄像系统及其校准方法和校准程序
US20130135474A1 (en) 2011-11-28 2013-05-30 Clarion Co., Ltd. Automotive Camera System and Its Calibration Method and Calibration Program
JP2013115540A (ja) 2011-11-28 2013-06-10 Clarion Co Ltd 車載カメラシステム及びその較正方法、及びその較正プログラム
CN103686143A (zh) 2012-09-24 2014-03-26 歌乐株式会社 照相机的校准方法以及装置
EP2711895A2 (en) 2012-09-24 2014-03-26 Clarion Co., Ltd. Calibration method and apparatus for in-vehicle camera
US20140085469A1 (en) * 2012-09-24 2014-03-27 Clarion Co., Ltd. Calibration Method and Apparatus for In-Vehicle Camera
WO2014054223A1 (ja) 2012-10-02 2014-04-10 株式会社デンソー キャリブレーション方法、及びキャリブレーション装置
JP2014074591A (ja) 2012-10-02 2014-04-24 Denso Corp キャリブレーション方法、及びキャリブレーション装置
US20150254853A1 (en) * 2012-10-02 2015-09-10 Denso Corporation Calibration method and calibration device
JP2014074632A (ja) 2012-10-03 2014-04-24 Isuzu Motors Ltd 車載ステレオカメラの校正装置及び校正方法

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
Chinese Office Action dated Sep. 4, 2018 in Chinese Application No. 201580020606.4.
Extended European Search Report issued in corresponding application No. 15 84 5604 dated Feb. 22, 2018.

Also Published As

Publication number Publication date
EP3203724A4 (en) 2018-03-28
US20170061622A1 (en) 2017-03-02
CN106233711B (zh) 2019-08-02
EP3203724A1 (en) 2017-08-09
EP3203724B1 (en) 2020-02-19
JP6371185B2 (ja) 2018-08-08
CN106233711A (zh) 2016-12-14
WO2016052505A1 (ja) 2016-04-07
JP2016070814A (ja) 2016-05-09

Similar Documents

Publication Publication Date Title
US10594943B2 (en) Camera calibration device and camera calibration system
EP3367677B1 (en) Calibration apparatus, calibration method, and calibration program
US9586455B2 (en) Road surface condition estimating apparatus
KR101787304B1 (ko) 교정 방법, 교정 장치 및 컴퓨터 프로그램 제품
JP6353289B2 (ja) 測距補正装置
EP2485203B1 (en) Vehicle-surroundings monitoring device
EP3262566A2 (en) Road vertical contour detection using a stabilized coordinate frame
JP2008182652A (ja) カメラ姿勢推定装置、車両、およびカメラ姿勢推定方法
JP2007261463A (ja) 車載カメラのキャリブレーションシステム
JP5959311B2 (ja) データ導出装置、及び、データ導出方法
CN109074480B (zh) 用于检测机动车辆的环境区域的图像中的滚动快门效应的方法、计算装置、驾驶员辅助系统以及机动车辆
US10083509B2 (en) Image calibration
WO2019073772A1 (ja) 移動体の位置推定装置及び位置推定方法
JP7280371B2 (ja) 車両コンポーネントの運動を補償するためのシステム、商用車両、方法およびコンピュータプログラム
JP2018136739A (ja) キャリブレーション装置
US11403770B2 (en) Road surface area detection device
KR20170114523A (ko) Avm용 주행 자동 공차 보정 장치 및 방법
JP7196920B2 (ja) 運転支援装置および運転支援方法、プログラム
JP6890293B2 (ja) カメラ校正装置、カメラ校正システム、カメラ校正方法及びプログラム
JP2018033035A (ja) 画像処理装置、撮影装置、プログラム、機器制御システム及び機器
KR20160097596A (ko) Avm 영상 보정 시스템 및 그 제어방법
JP2020087210A (ja) キャリブレーション装置及びキャリブレーション方法
JP2013005035A (ja) 車載カメラ姿勢検出装置

Legal Events

Date Code Title Description
AS Assignment

Owner name: CLARION CO., LTD., JAPAN

Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:SAKANO, MORIHIKO;SATO, KEIJI;REEL/FRAME:040122/0181

Effective date: 20160912

STPP Information on status: patent application and granting procedure in general

Free format text: NON FINAL ACTION MAILED

STPP Information on status: patent application and granting procedure in general

Free format text: RESPONSE TO NON-FINAL OFFICE ACTION ENTERED AND FORWARDED TO EXAMINER

STPP Information on status: patent application and granting procedure in general

Free format text: NOTICE OF ALLOWANCE MAILED -- APPLICATION RECEIVED IN OFFICE OF PUBLICATIONS

STPP Information on status: patent application and granting procedure in general

Free format text: PUBLICATIONS -- ISSUE FEE PAYMENT VERIFIED

STCF Information on status: patent grant

Free format text: PATENTED CASE

MAFP Maintenance fee payment

Free format text: PAYMENT OF MAINTENANCE FEE, 4TH YEAR, LARGE ENTITY (ORIGINAL EVENT CODE: M1551); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY

Year of fee payment: 4